| *** kevinc_ has quit IRC | 00:01 | |
| *** sileht has quit IRC | 00:06 | |
| *** kenhui has joined #openstack-swift | 00:12 | |
| *** lpabon has quit IRC | 00:19 | |
| *** lpabon has joined #openstack-swift | 00:20 | |
| notmyname | anyone here who know's about sphinx? | 00:20 |
|---|---|---|
| torgomatic | if you get its riddles wrong, it eats you | 00:20 |
| notmyname | I'm being eaten. It's a riddle to figure out how to reference files not in the docs/source dir | 00:21 |
| *** dmorita has joined #openstack-swift | 00:22 | |
| *** lpabon has quit IRC | 00:24 | |
| *** matsuhashi has joined #openstack-swift | 00:24 | |
| *** jamiehannaford has quit IRC | 00:24 | |
| notmyname | acoles: portante: I'm just about done with getting swift-specs set up. fighting one sphinx error now | 00:25 |
| notmyname | acoles: FYI, you'll have to do more than rebase your change. the dir structure I'm setting up will mean your spec file needs to move | 00:25 |
| notmyname | (it will be obvious to where) | 00:25 |
| *** mwstorer has quit IRC | 00:31 | |
| zaitcev | the problem is, once the manual is processed, it's not in docs/source dir. You can always refer to file:///, but... | 00:34 |
| notmyname | zaitcev: so it needs to be in both locations? | 00:35 |
| notmyname | zaitcev: specifically in this case it's the project README.rst file and others like that | 00:35 |
| zaitcev | notmyname: I mean that manual is supposed to be delivered somewhere, deployed on a server, etc. It makes no sense for it to refer to a file. Someone may be reading it on an iPad. So... Refer to http://gitrub.com/openstack/swift/README.rst at best. | 00:36 |
| notmyname | hmm | 00:36 |
| *** kota_ has joined #openstack-swift | 00:37 | |
| notmyname | that makes sense for the final build/html/* files, but not for the `python ./setup.py build_sphinx` step | 00:37 |
| notmyname | eg in swift-specs, it would be nice that the generated docs include the contents of README, CONTRIBUTING, and the example template | 00:38 |
| zaitcev | I see | 00:38 |
| notmyname | build_sphinx should pull that in to build/html | 00:38 |
| notmyname | which can be shipped around | 00:38 |
| zaitcev | right, makes sense | 00:38 |
| notmyname | but I can't find how to reference those high-level files (And it seems weird to include them twice) | 00:38 |
| mattoliverau | notmyname: I don't know sphinx but with some google foo can't you just use the reletive path and the include? http://reinout.vanrees.org/weblog/2010/12/08/include-external-in-sphinx.html or there is this method http://stackoverflow.com/questions/10199233/can-sphinx-link-to-documents-that-are-not-located-in-directories-below-the-root (the non-symlink answer) | 00:41 |
| notmyname | mattoliverau: it seems you have better google-fu than me ;-) | 00:43 |
| *** kenhui1 has joined #openstack-swift | 00:43 | |
| mattoliverau | notmyname: lol, the google-fu is string with this one :P | 00:44 |
| notmyname | string, yarn. whatever ;-) | 00:44 |
| *** kenhui has quit IRC | 00:44 | |
| *** shri has left #openstack-swift | 00:45 | |
| *** kopparam has joined #openstack-swift | 00:49 | |
| notmyname | mattoliverau: it works!!! thanks | 00:49 |
| mattoliverau | woo! | 00:49 |
| openstackgerrit | John Dickinson proposed a change to openstack/swift-specs: Add the initial specs framework https://review.openstack.org/102417 | 00:51 |
| notmyname | mordred: FYI ^^ | 00:52 |
| zaitcev | but you added README.rst copy | 00:53 |
| zaitcev | ignore that | 00:53 |
| openstackgerrit | Samuel Merritt proposed a change to openstack/swift: Zero-copy object-server GET responses with splice() https://review.openstack.org/102609 | 00:53 |
| notmyname | acoles: portante: clayg: so that's the proposed swift-specs repo. check it out, especially the structure in README | 00:53 |
| notmyname | zaitcev: :-) | 00:53 |
| *** kopparam has quit IRC | 00:53 | |
| *** baojg has joined #openstack-swift | 00:53 | |
| notmyname | (I'd love comments from everyone, just those 3 have been bugging me about it) | 00:53 |
| *** baojg_ has joined #openstack-swift | 01:01 | |
| *** baojg has quit IRC | 01:02 | |
| *** baojg_ has quit IRC | 01:05 | |
| *** kenhui1 has quit IRC | 01:17 | |
| *** mkollaro has quit IRC | 01:25 | |
| anticw | does anyone here know if swift.common.internal_client is overly robust? | 01:35 |
| anticw | i'm asking because delete_object throws exceptions when it encouters a 404 ... which isn't something i expect or can see why at a glance from the code | 01:35 |
| anticw | what's more ... it throws urllib2.HTTPError not .ClientException ... | 01:36 |
| *** nosnos has joined #openstack-swift | 01:47 | |
| *** mlipchuk has quit IRC | 02:11 | |
| *** nsquare has quit IRC | 02:21 | |
| *** bkopilov has quit IRC | 02:24 | |
| *** dmsimard_away is now known as dmsimard | 02:34 | |
| *** cm2 has quit IRC | 02:37 | |
| *** baojg has joined #openstack-swift | 02:45 | |
| *** kopparam has joined #openstack-swift | 02:52 | |
| *** kopparam has quit IRC | 02:57 | |
| *** mhu has quit IRC | 03:05 | |
| *** gvernik has joined #openstack-swift | 03:05 | |
| *** jyoti-ranjan has joined #openstack-swift | 03:07 | |
| yuan | zaitcev: met some PasteDepoly issue on CentOS6.5. w/ PasteDepoly 1.5.2, swift refuses to start: Error trying to load config from /etc/swift/proxy-server.conf: pastedeploy>=1.3.3. but w/ PasteDeploy 1.5.0 everything seems OK | 03:08 |
| *** mhu has joined #openstack-swift | 03:10 | |
| *** gvernik has quit IRC | 03:11 | |
| zaitcev | yuan: try to download http://people.redhat.com/zaitcev/tmp/xattr_load.py and run python xattr_load.py| grep -i paste | 03:24 |
| zaitcev | yuan: honestly I have no idea what the problem may be, especially since we don't know where the offending PasteDeploy originates, if you corrupted anything with pip, etc. | 03:26 |
| zaitcev | Considering that F20 has python-paste-deploy-1.5.0-10.fc20.noarch, I would suspect it's something CentOS people screwed up, or you picked it at any other rogue repo. | 03:27 |
| zaitcev | (or worse, installed with pip) | 03:27 |
| yuan | oh yes, i used 'pip install -r requirements' | 03:28 |
| zaitcev | so, what does the command above say? | 03:29 |
| zaitcev | on my system it's | 03:29 |
| zaitcev | PasteDeploy(pastedeploy) 1.5.0 /usr/lib/python2.7/site-packages | 03:29 |
| zaitcev | Paste(paste) 1.7.5.1 /usr/lib/python2.7/site-packages | 03:29 |
| yuan | no outputs but pip list can see PasteDeploy | 03:33 |
| zaitcev | That doesn't count, obviously, because everyone else's loadapp() ends in pkg_resources. If they can't find what they need, the can't import anything. | 03:36 |
| zaitcev | Well, they probably could if they just... I dunno... used the right path or something. | 03:37 |
| yuan | hmm, should be something wrong with pip. let have a try again. thanks a lot | 03:37 |
| zaitcev | np. Sadly I don't understand how the whole machinery works. | 03:38 |
| portante | notmyname: looking ... | 03:38 |
| zaitcev | But you can keep re-running the script I wrote and tweak things until it works. | 03:38 |
| *** Kbee has joined #openstack-swift | 03:40 | |
| portante | just found out that you can add comments in gerrit to a whole file, not just a specific line | 03:42 |
| anticw | zaitcev: RHEL 6.x has python 2.7 available? | 03:47 |
| *** dmsimard is now known as dmsimard_away | 03:47 | |
| zaitcev | anticw: It should... RHEL 6 GA shipped with 2.6, but I think it moved to 2.7. | 03:48 |
| zaitcev | go go gadget Koji | 03:48 |
| anticw | yum would find/install that right? and if i don't see it then it's because i starting drinking earlier tonight? | 03:48 |
| *** jkelly has joined #openstack-swift | 03:49 | |
| zaitcev | hmm | 03:50 |
| *** baojg has quit IRC | 03:50 | |
| zaitcev | looks like it's still 2.6, at python-2.6.6-53.el6 currently | 03:50 |
| *** baojg has joined #openstack-swift | 03:50 | |
| zaitcev | Last build was in January | 03:51 |
| anticw | upgrade to rhel7! | 03:53 |
| anticw | zaitcev: actually, i looked at rdo+selinux ... and comments there on what selinux brings to the table? | 03:53 |
| *** kopparam has joined #openstack-swift | 03:53 | |
| *** matsuhashi has quit IRC | 03:54 | |
| *** zhiyan_ is now known as zhiyan | 03:55 | |
| *** baojg has quit IRC | 03:55 | |
| *** matsuhashi has joined #openstack-swift | 03:55 | |
| *** kopparam has quit IRC | 03:58 | |
| *** bkopilov has joined #openstack-swift | 03:59 | |
| *** matsuhashi has quit IRC | 03:59 | |
| *** Italux has joined #openstack-swift | 04:02 | |
| *** Italux has quit IRC | 04:02 | |
| *** jonkelly has joined #openstack-swift | 04:05 | |
| *** jkelly has quit IRC | 04:05 | |
| *** jonkelly is now known as jkelly | 04:05 | |
| *** baojg has joined #openstack-swift | 04:08 | |
| *** baojg has quit IRC | 04:08 | |
| *** baojg has joined #openstack-swift | 04:08 | |
| *** nosnos has quit IRC | 04:09 | |
| *** jyoti-ranjan has quit IRC | 04:22 | |
| *** nshaikh has joined #openstack-swift | 04:27 | |
| zaitcev | anticw: I think SElinux does very little for proxy and nothing for a/c/o. We didn't have a proper remote exploit in years in Swift. | 04:29 |
| *** jyoti-ranjan has joined #openstack-swift | 04:29 | |
| zaitcev | anticw: the whole value is in need-based access, which mitigates security issues in apps. Very valuable for something like Firefox. | 04:30 |
| *** matsuhashi has joined #openstack-swift | 04:37 | |
| *** nosnos has joined #openstack-swift | 04:37 | |
| *** kopparam has joined #openstack-swift | 04:44 | |
| jkelly | it’s more useful where you are running many apps with different roles on the same server… basically its value is predicated on the fact that once a service on the system has been compromised that that service can be leveraged to access other data on the same system | 04:53 |
| *** kopparam has quit IRC | 04:54 | |
| jkelly | in a ‘cloudy’ system where servers/vms are more single-purpose, much of the value is eliminated | 04:54 |
| jkelly | one of the primary problems is that its utility is predicated upon a service being compromised in the first place… typically you are better of focusing on your front-line (firewall rules, network security, patching strategy) before worrying about something like SELinux. Not to say there’s no value there, but I do question its off-the-shelf utility as RedHat implements it | 04:55 |
| *** ajc_ has joined #openstack-swift | 04:59 | |
| *** psharma has joined #openstack-swift | 05:00 | |
| *** ppai has joined #openstack-swift | 05:10 | |
| *** ppai has quit IRC | 05:15 | |
| *** kopparam has joined #openstack-swift | 05:16 | |
| *** kopparam_ has joined #openstack-swift | 05:19 | |
| *** kopparam has quit IRC | 05:19 | |
| *** zaitcev has quit IRC | 05:20 | |
| *** jkelly has quit IRC | 05:22 | |
| *** gyee has quit IRC | 05:22 | |
| *** ppai has joined #openstack-swift | 05:28 | |
| *** kopparam_ has quit IRC | 05:33 | |
| *** kopparam has joined #openstack-swift | 05:33 | |
| *** kopparam has quit IRC | 05:40 | |
| *** kopparam has joined #openstack-swift | 05:44 | |
| *** jyoti-ranjan has quit IRC | 05:55 | |
| *** kota_ has quit IRC | 05:56 | |
| *** Kbee has quit IRC | 05:57 | |
| *** Kbee has joined #openstack-swift | 05:57 | |
| *** jyoti-ranjan has joined #openstack-swift | 06:00 | |
| *** jyoti-ranjan has quit IRC | 06:03 | |
| *** jyoti-ranjan has joined #openstack-swift | 06:06 | |
| *** jyoti-ranjan has quit IRC | 06:07 | |
| *** jyoti-ranjan has joined #openstack-swift | 06:09 | |
| *** matsuhashi has quit IRC | 06:13 | |
| *** zhiyan is now known as zhiyan_ | 06:13 | |
| *** matsuhashi has joined #openstack-swift | 06:13 | |
| *** zhiyan_ is now known as zhiyan | 06:14 | |
| *** matsuhashi has quit IRC | 06:18 | |
| *** matsuhashi has joined #openstack-swift | 06:18 | |
| *** jyoti-ranjan has quit IRC | 06:18 | |
| *** jyoti-ranjan has joined #openstack-swift | 06:23 | |
| *** matsuhashi has quit IRC | 06:24 | |
| *** matsuhashi has joined #openstack-swift | 06:24 | |
| *** kopparam has quit IRC | 06:30 | |
| *** matsuhashi has quit IRC | 06:31 | |
| *** matsuhashi has joined #openstack-swift | 06:32 | |
| *** baojg has quit IRC | 06:34 | |
| *** matsuhas_ has joined #openstack-swift | 06:34 | |
| goodes | peluse_: any hints where I can find specs for erasure coding support? | 06:35 |
| *** baojg has joined #openstack-swift | 06:36 | |
| *** matsuhashi has quit IRC | 06:37 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fix double request in simple client https://review.openstack.org/102746 | 06:42 |
| *** sileht has joined #openstack-swift | 06:44 | |
| *** chandan_kumar has joined #openstack-swift | 06:47 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fix KeyError on Storage Policy Container Sync https://review.openstack.org/102747 | 06:55 |
| *** tsg has quit IRC | 06:56 | |
| *** jamiehannaford has joined #openstack-swift | 06:57 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fix KeyError on Storage Policy Container Sync https://review.openstack.org/102747 | 06:58 |
| *** haomaiwa_ has joined #openstack-swift | 07:04 | |
| bkopilov | Hi , Question about swift cross_domain | 07:04 |
| bkopilov | I am running automation and one of the swift tests fails ( tempest upstream code ) | 07:04 |
| bkopilov | fails on get_crossdomain_policy | 07:05 |
| *** openstackgerrit has quit IRC | 07:07 | |
| *** haomaiw__ has joined #openstack-swift | 07:08 | |
| *** haomaiwa_ has quit IRC | 07:08 | |
| *** kopparam has joined #openstack-swift | 07:11 | |
| *** mmcardle has joined #openstack-swift | 07:14 | |
| *** kopparam has quit IRC | 07:15 | |
| *** kopparam has joined #openstack-swift | 07:16 | |
| *** kopparam has quit IRC | 07:16 | |
| *** jamiehannaford has quit IRC | 07:17 | |
| *** jamiehannaford has joined #openstack-swift | 07:18 | |
| *** foexle has joined #openstack-swift | 07:19 | |
| *** jamiehan_ has joined #openstack-swift | 07:26 | |
| *** jamiehannaford has quit IRC | 07:30 | |
| *** eglynn has quit IRC | 07:31 | |
| *** jamiehan_ has quit IRC | 07:33 | |
| *** jamiehannaford has joined #openstack-swift | 07:39 | |
| *** mlipchuk has joined #openstack-swift | 07:44 | |
| *** Kbee has quit IRC | 07:45 | |
| *** Kbee has joined #openstack-swift | 07:45 | |
| *** kopparam has joined #openstack-swift | 07:47 | |
| *** pconstantine_ has joined #openstack-swift | 07:51 | |
| *** kopparam has quit IRC | 07:52 | |
| *** nsquare has joined #openstack-swift | 07:55 | |
| *** joeljwright has joined #openstack-swift | 07:58 | |
| *** nsquare_ has joined #openstack-swift | 08:01 | |
| *** nsquare has quit IRC | 08:04 | |
| *** nacim has joined #openstack-swift | 08:05 | |
| *** eglynn has joined #openstack-swift | 08:20 | |
| cschwede | acoles: Good Morning Alistair! I reviewed your Keystone V3 swiftclient patch (https://review.openstack.org/#/c/91788/), looks good to me so far, just a minor nitpick. Let me know what you think about it | 08:22 |
| *** kopparam has joined #openstack-swift | 08:23 | |
| *** jyoti-ranjan has quit IRC | 08:30 | |
| *** jyoti-ranjan has joined #openstack-swift | 08:39 | |
| *** matsuhas_ has quit IRC | 08:56 | |
| *** matsuhashi has joined #openstack-swift | 08:57 | |
| *** matsuhashi has quit IRC | 09:01 | |
| *** matsuhashi has joined #openstack-swift | 09:01 | |
| *** matsuhashi has quit IRC | 09:07 | |
| *** matsuhashi has joined #openstack-swift | 09:08 | |
| *** matsuhashi has quit IRC | 09:08 | |
| *** matsuhashi has joined #openstack-swift | 09:09 | |
| *** pconstantine_ has quit IRC | 09:10 | |
| *** mkerrin has quit IRC | 09:18 | |
| *** kopparam has quit IRC | 09:22 | |
| *** mkerrin has joined #openstack-swift | 09:25 | |
| *** dANOKELOFF has joined #openstack-swift | 09:37 | |
| *** kopparam has joined #openstack-swift | 09:37 | |
| *** dmorita has quit IRC | 09:40 | |
| *** nosnos has quit IRC | 09:52 | |
| *** baojg has quit IRC | 10:05 | |
| *** baojg has joined #openstack-swift | 10:06 | |
| *** matsuhashi has quit IRC | 10:09 | |
| *** mkollaro has joined #openstack-swift | 10:09 | |
| *** matsuhashi has joined #openstack-swift | 10:09 | |
| *** matsuhashi has quit IRC | 10:10 | |
| *** matsuhashi has joined #openstack-swift | 10:11 | |
| *** mkollaro has quit IRC | 10:14 | |
| *** mkollaro has joined #openstack-swift | 10:15 | |
| *** matsuhashi has quit IRC | 10:16 | |
| *** matsuhashi has joined #openstack-swift | 10:17 | |
| *** matsuhas_ has joined #openstack-swift | 10:20 | |
| *** matsuhashi has quit IRC | 10:21 | |
| *** kopparam has quit IRC | 10:25 | |
| *** matsuhas_ has quit IRC | 10:26 | |
| *** matsuhashi has joined #openstack-swift | 10:26 | |
| *** matsuhashi has quit IRC | 10:35 | |
| *** matsuhashi has joined #openstack-swift | 10:35 | |
| *** matsuhashi has quit IRC | 10:35 | |
| *** matsuhashi has joined #openstack-swift | 10:36 | |
| *** dANOKELOFF has quit IRC | 10:37 | |
| *** Kbee has quit IRC | 10:48 | |
| *** chandan_kumar is now known as chkumar246 | 10:51 | |
| *** chkumar246 is now known as chandan_kumar | 10:53 | |
| *** chandan_kumar has quit IRC | 10:54 | |
| *** chandan_kumar has joined #openstack-swift | 10:54 | |
| *** kopparam has joined #openstack-swift | 10:56 | |
| *** kopparam has quit IRC | 11:03 | |
| *** matsuhashi has quit IRC | 11:05 | |
| *** matsuhashi has joined #openstack-swift | 11:05 | |
| *** haomaiw__ has quit IRC | 11:05 | |
| *** haomaiwang has joined #openstack-swift | 11:06 | |
| *** kopparam has joined #openstack-swift | 11:09 | |
| *** matsuhashi has quit IRC | 11:10 | |
| *** nsquare_ has quit IRC | 11:14 | |
| *** nsquare has joined #openstack-swift | 11:17 | |
| *** cm2 has joined #openstack-swift | 11:22 | |
| *** nsquare has quit IRC | 11:22 | |
| *** dANOKELOFF has joined #openstack-swift | 11:30 | |
| *** haomaiwang has quit IRC | 11:58 | |
| *** marcusvrn has joined #openstack-swift | 11:59 | |
| *** ppai has quit IRC | 12:02 | |
| *** baojg has quit IRC | 12:02 | |
| *** baojg has joined #openstack-swift | 12:03 | |
| *** mmcardle has quit IRC | 12:03 | |
| *** baojg_ has joined #openstack-swift | 12:05 | |
| *** baojg_ has quit IRC | 12:05 | |
| *** baojg_ has joined #openstack-swift | 12:06 | |
| *** baojg has quit IRC | 12:06 | |
| *** baojg_ has quit IRC | 12:10 | |
| *** ppai has joined #openstack-swift | 12:18 | |
| *** mmcardle has joined #openstack-swift | 12:21 | |
| *** kopparam has quit IRC | 12:21 | |
| jyoti-ranjan | I have question related to Swift namespace of region and zone. | 12:23 |
| jyoti-ranjan | As per my understanding, Swift region and zone are different from nova availability zone and keystone region. Am I correct? | 12:24 |
| jyoti-ranjan | Certainly, we can leverage nova availability zone to mape to Swift zone but I believe that these nsamespaces are different. | 12:24 |
| *** openstackgerrit has joined #openstack-swift | 12:26 | |
| cschwede | jyoti-ranjan: yes, these are different - in fact Swift regions and zones are completely independet and configured in the rings | 12:29 |
| cschwede | s/independet/independent/ | 12:29 |
| jyoti-ranjan | I was just having idea of mapping nova availability zone to Swift zone. I was thinking of flow like this: | 12:30 |
| jyoti-ranjan | 1) User creates 3 nova availability zone for Swift | 12:30 |
| jyoti-ranjan | 2) Swift goes and provision node from that zone and expand ring if required | 12:31 |
| jyoti-ranjan | For provisioning, we can use Ironic or something | 12:31 |
| jyoti-ranjan | Does it make sense to do something like this as it will assist operator to use nova tags itself to define Swift zones | 12:32 |
| jyoti-ranjan | Certainly, it is not a good idea if customer wants to deploy only Swift | 12:32 |
| jyoti-ranjan | But many installation has nova as well. | 12:33 |
| jyoti-ranjan | Any thoughts? | 12:33 |
| *** kopparam has joined #openstack-swift | 12:33 | |
| *** bkopilov has quit IRC | 12:34 | |
| *** tdasilva has joined #openstack-swift | 12:34 | |
| cschwede | 2) would not be done by Swift; in fact all ring building is currently done either manually or by tools (puppet, chef, ansible, …). | 12:39 |
| cschwede | thus i think this is more a task outside of Swift - making it suitable for deployments with or without Nova | 12:40 |
| cschwede | so where do you want to have this tool? In Horizon, or on the command line? | 12:41 |
| cschwede | jyoti-ranjan: ^^ | 12:41 |
| *** mkollaro1 has joined #openstack-swift | 12:43 | |
| *** mkollaro has quit IRC | 12:43 | |
| *** annegent_ has joined #openstack-swift | 12:44 | |
| *** annegent_ is now known as annegentle_ | 12:44 | |
| *** eglynn is now known as eglynn-lunch-afk | 12:49 | |
| *** eglynn-lunch-afk has quit IRC | 12:54 | |
| *** baojg has joined #openstack-swift | 12:56 | |
| *** ajc_ has quit IRC | 13:00 | |
| *** Kbee has joined #openstack-swift | 13:05 | |
| *** annegentle_ has quit IRC | 13:06 | |
| *** ppai has quit IRC | 13:06 | |
| jyoti-ranjan | <cschwede>We wanted to run that tool in separate node | 13:11 |
| *** miqui has joined #openstack-swift | 13:11 | |
| jyoti-ranjan | calling it as Swift controller | 13:11 |
| jyoti-ranjan | Yes, you are very right that provisioning zone nodes can not be scopt of Swift. I wanted your advice on whether marrying nova availability zone with Swift zone is good idea or not. | 13:12 |
| cschwede | jyoti-ranjan: i think most deployments will use different zones for Nova and Swift, because you need different kind of hardware for the zones | 13:13 |
| cschwede | and also different kind of backend network traffic | 13:13 |
| cschwede | so i think mapping a Nova zone to a Swift zone is a special case, most likely they are different | 13:13 |
| jyoti-ranjan | I see | 13:14 |
| jyoti-ranjan | It makes sense | 13:14 |
| jyoti-ranjan | <cschwede> Thank you very much. | 13:19 |
| cschwede | jyoti-ranjan: you’re welcome, let me know if you have more questions! | 13:19 |
| *** kopparam has quit IRC | 13:25 | |
| *** kopparam has joined #openstack-swift | 13:25 | |
| *** zz_wasmum is now known as wasmum | 13:27 | |
| *** rook_ has joined #openstack-swift | 13:28 | |
| *** kopparam has quit IRC | 13:30 | |
| *** shausy has joined #openstack-swift | 13:30 | |
| peluse_ | goodes: There's some design notes attached to one of the discussions on our discussion board: https://trello.com/c/t1lud5iA/3-ec-reconstructor-general-design-discussion | 13:36 |
| * peluse_ forgot to mention yesterday - on vacation Wed-Fri this week | 13:38 | |
| peluse_ | goodes: let me know if you have specific questions, others can help too. I'll be more offline than online for the next two days though :) | 13:38 |
| *** joeljwright has left #openstack-swift | 13:38 | |
| *** psharma has quit IRC | 13:40 | |
| rook_ | administering swift question for you guys: would it "hurt" anything to set all the files in /etc/swift to root:swift ownership and 0640 permission ? | 13:41 |
| *** eglynn-lunch-afk has joined #openstack-swift | 13:46 | |
| *** mkollaro has joined #openstack-swift | 13:46 | |
| *** mkollaro1 has quit IRC | 13:50 | |
| *** annegent_ has joined #openstack-swift | 13:55 | |
| *** jasond has left #openstack-swift | 13:56 | |
| *** dmsimard_away is now known as dmsimard | 14:00 | |
| *** annegent_ has quit IRC | 14:01 | |
| *** dmsimard is now known as dmsimard_away | 14:02 | |
| *** nacim has quit IRC | 14:04 | |
| *** zaitcev has joined #openstack-swift | 14:06 | |
| *** ChanServ sets mode: +v zaitcev | 14:06 | |
| *** annegent_ has joined #openstack-swift | 14:06 | |
| *** dmsimard_away has quit IRC | 14:08 | |
| *** dmsimard has joined #openstack-swift | 14:11 | |
| *** kenhui has joined #openstack-swift | 14:19 | |
| *** eglynn-lunch-afk is now known as eglynn | 14:22 | |
| *** mhu has quit IRC | 14:27 | |
| *** mhu has joined #openstack-swift | 14:28 | |
| wasmum | good morning | 14:30 |
| wasmum | anyone have some swift3 keystone time? | 14:30 |
| *** jamiehannaford has quit IRC | 14:36 | |
| *** baojg has quit IRC | 14:39 | |
| peluse_ | goodes: also, there's a lot more EC specific infomation at https://bitbucket.org/kmgreen2/pyeclib where the library that we'll be using is being developed | 14:42 |
| *** nacim has joined #openstack-swift | 14:50 | |
| *** annegent_ has quit IRC | 14:52 | |
| *** annegent_ has joined #openstack-swift | 14:53 | |
| *** bkopilov has joined #openstack-swift | 14:53 | |
| *** tongli has joined #openstack-swift | 14:56 | |
| *** Kbee has quit IRC | 14:59 | |
| *** nshaikh has left #openstack-swift | 15:02 | |
| *** nshaikh has joined #openstack-swift | 15:07 | |
| *** annegent_ has quit IRC | 15:08 | |
| *** kevinc_ has joined #openstack-swift | 15:12 | |
| *** shausy has quit IRC | 15:24 | |
| *** JelleB is now known as a1|away | 15:26 | |
| *** lpabon has joined #openstack-swift | 15:29 | |
| *** annegent_ has joined #openstack-swift | 15:30 | |
| *** nacim has quit IRC | 15:30 | |
| *** mkollaro has quit IRC | 15:34 | |
| *** lpabon has quit IRC | 15:39 | |
| *** mkollaro has joined #openstack-swift | 15:40 | |
| *** annegent_ has quit IRC | 15:40 | |
| *** kopparam has joined #openstack-swift | 15:41 | |
| *** nacim has joined #openstack-swift | 15:43 | |
| *** jyoti-ranjan has quit IRC | 15:43 | |
| *** lpabon has joined #openstack-swift | 15:48 | |
| *** lpabon has quit IRC | 15:49 | |
| *** lpabon has joined #openstack-swift | 15:53 | |
| clayg | rook_: do you run processes as root or swift? should be fine, nothing gets written into /etc/swift afaik. | 15:54 |
| *** kopparam has quit IRC | 15:55 | |
| *** kopparam has joined #openstack-swift | 15:55 | |
| rook_ | clayg: I thought it would be fine too, however, when the workers fork under the swift user they can't read their conf file (even though they are group readable) -- strange | 15:55 |
| clayg | rook_: some instructions put the .builder files in /etc/swift and to write out new .ring.gz's if you're in /etc/swift looks like you'd have to run swift-ring-builder as root - but you should really be managing your .builder's on a seperate box and pushing the ring.gz's out - so as long as they end up readable it should be fine.. | 15:56 |
| clayg | awww man i thought all that permissions drop group shiz got fixed - what version are you running? | 15:57 |
| rook_ | 1.10 | 15:57 |
| clayg | so the commit that fixed that was c656e18; oh there's a bug #1269473 | 15:59 |
| *** bvandenh has joined #openstack-swift | 15:59 | |
| clayg | looks like that would be in ~1.12 | 15:59 |
| *** annegent_ has joined #openstack-swift | 15:59 | |
| clayg | which is not as long ago as I thought, current is 1.13 with 2.0 rc out | 16:00 |
| *** kopparam has quit IRC | 16:00 | |
| *** chandan_kumar has quit IRC | 16:02 | |
| *** bvandenh has quit IRC | 16:04 | |
| clayg | rook_: ps no fair asking "would this work" if you've already tested it :P | 16:04 |
| zaitcev | I didn't notice that regression because I scp *.ring.gz | 16:05 |
| *** gyee has joined #openstack-swift | 16:05 | |
| *** mwstorer has joined #openstack-swift | 16:11 | |
| *** gyee has quit IRC | 16:11 | |
| *** dANOKELOFF has quit IRC | 16:14 | |
| *** annegent_ has quit IRC | 16:15 | |
| *** annegent_ has joined #openstack-swift | 16:16 | |
| *** kenhui has quit IRC | 16:19 | |
| *** gyee has joined #openstack-swift | 16:20 | |
| *** lpabon_test has joined #openstack-swift | 16:23 | |
| *** kenhui has joined #openstack-swift | 16:24 | |
| *** kevinc_ has quit IRC | 16:31 | |
| *** kevinc_ has joined #openstack-swift | 16:31 | |
| *** zhiyan is now known as zhiyan_ | 16:32 | |
| *** nacim has quit IRC | 16:34 | |
| goodes | peluse_: I was wondering regarding meta data for ec objects - is the plan to store them with xattr with all the parts or store the meta data out-of-band in a seperate file? | 16:40 |
| rook_ | clayg: :) | 16:41 |
| *** lpabon_test has quit IRC | 16:47 | |
| *** pberis has joined #openstack-swift | 16:48 | |
| *** mmcardle has quit IRC | 16:48 | |
| goodes | peluse_: i.e. will the performance of updating metadata on ec vs ring objects being slower/same/faster? | 16:55 |
| *** AdnanCloud1 has joined #openstack-swift | 17:00 | |
| AdnanCloud1 | hello! | 17:01 |
| AdnanCloud1 | this the right room for swift programming? | 17:01 |
| cschwede | AdnanCloud1: Hello! Programming for Openstack Swift (the object storage) or Apple Swift? | 17:01 |
| peluse_ | goodes: there's no change to how obj metadata is managed wrt EC; there are a few additional items that are stored in xattrs though. Details of the current plan are in that trello card attachment I mention ealier | 17:01 |
| AdnanCloud1 | apple swift | 17:02 |
| AdnanCloud1 | openstack swift.. never heard of this actually.. what is it | 17:03 |
| cschwede | AdnanCloud1: No, this is for Openstack Swift, the object storage system. | 17:03 |
| AdnanCloud1 | where is this used | 17:03 |
| AdnanCloud1 | amazon? | 17:03 |
| cschwede | AdnanCloud1: HP, Rackspacke, Disney, Cloudwatt, to name just a few big installations | 17:03 |
| AdnanCloud1 | right okay..hmm let me google this | 17:04 |
| cschwede | AdnanCloud1: http://docs.openstack.org/developer/swift/ | 17:04 |
| *** lpabon has quit IRC | 17:04 | |
| AdnanCloud1 | okay looks interesting... but i can only learn so much :) this will have to wait for another time | 17:05 |
| *** lpabon has joined #openstack-swift | 17:06 | |
| Alex_Gaynor | cschwede: Don't forget Wikipedia! | 17:06 |
| cschwede | Alex_Gaynor: oh yes, probably one of the most widely used installations! | 17:07 |
| notmyname | AdnanCloud1: good morning (just got out of a meeting) | 17:08 |
| notmyname | AdnanCloud1: I think the apple swift language stuff is #swift-lang (or maybe that's the other swift programming language) | 17:08 |
| notmyname | AdnanCloud1: this swift that we work on--openstack swift--is a storage system. you can think of it like "S3 in your own datacenter" | 17:08 |
| *** AdnanCloud1 has quit IRC | 17:09 | |
| *** AdnanCloud1 has joined #openstack-swift | 17:09 | |
| chmouel | we *should* really implement openstack swift in swift to make it not confusing | 17:09 |
| notmyname | chmouel: right ;-) | 17:10 |
| *** annegent_ has quit IRC | 17:12 | |
| chmouel | cschwede: nice goal :) | 17:13 |
| *** annegent_ has joined #openstack-swift | 17:13 | |
| *** annegent_ has quit IRC | 17:18 | |
| * peluse_ out for most of the rest of the day.... | 17:19 | |
| notmyname | portante: thanks for looking at the specs repo proposal | 17:23 |
| portante | notmyname: welcome | 17:24 |
| notmyname | clayg: have you taken a look at the specs repo proposal yet? how much do you care? | 17:24 |
| notmyname | portante: you think the proposed repo structure makes sense? | 17:24 |
| portante | seems fine to be to start. I think we should get it in place and let folks starting working with it so we can tweak it sooner than later | 17:25 |
| notmyname | ya. I definitely think it's a living thing | 17:25 |
| openstackgerrit | Joel Wright proposed a change to openstack/python-swiftclient: Add importable SwiftService incorporating shell.py logic https://review.openstack.org/85453 | 17:27 |
| *** kenhui has quit IRC | 17:29 | |
| *** kevinc_ has quit IRC | 17:32 | |
| *** kenhui has joined #openstack-swift | 17:38 | |
| *** kevinc_ has joined #openstack-swift | 17:39 | |
| clayg | notmyname: i'm not sure exactly what i should be looking for? I commit .rst docs when I want design feedback - seems stright forward. I really just want to look at acoles docs for fast-post | 17:39 |
| notmyname | clayg: the proposed structure in the README is IMO the only real thing. and that the generate sphinx docs actually work (they do, I checked) | 17:40 |
| *** mlipchuk has quit IRC | 17:40 | |
| *** chandan_kumar has joined #openstack-swift | 17:41 | |
| *** nsquare has joined #openstack-swift | 17:42 | |
| *** shri has joined #openstack-swift | 17:42 | |
| *** pberis has quit IRC | 17:45 | |
| *** annegent_ has joined #openstack-swift | 17:55 | |
| notmyname | clayg: if you're ok with it (or don't care too much yet), then I'll go ahead an land it. if you want changes, let me know | 17:56 |
| clayg | notmyname: the .gitreview change s/object/swift is sorta annoying | 17:58 |
| notmyname | clayg: that it's called swift-specs instead of object-specs? | 17:58 |
| *** mmcardle has joined #openstack-swift | 17:59 | |
| *** annegent_ has quit IRC | 18:00 | |
| clayg | notmyname: running tox to build docs the first time fails -> https://gist.github.com/clayg/a5317b40563a5931e16e | 18:00 |
| notmyname | clayg: that's the pattern being adopted across openstack. eg nova-specs, keystone-specs, etc. not compute-specs, bare-metal-deployment-specs, identity-management-specs, etc | 18:00 |
| notmyname | clayg: and works the second time? | 18:01 |
| clayg | notmyname: yup, you can make it fail by 'rm -fr doc/build' | 18:01 |
| clayg | notmyname: run it twice and it's good to go | 18:01 |
| clayg | mordred: ^ dunno why, probably not a big deal | 18:02 |
| notmyname | clayg: WOMM: https://gist.github.com/notmyname/b313913998bf2913ad2d | 18:02 |
| clayg | notmyname: yeah so i normally checkout from github and then git review and the one in github has object-specs so i had to fix it by hand (and remove the broken gerrit remtoe) to make git review work | 18:02 |
| clayg | notmyname: weird, git clean -fd too maybe? | 18:03 |
| notmyname | nope. works | 18:03 |
| clayg | notmyname: maybe i broke something trying to rebase acoles change onto the initial? | 18:04 |
| notmyname | maybe a requirements thing. I removed some requirements as I was working on it but maybe got something installed? | 18:04 |
| notmyname | still works if I `rm -rf .tox/` | 18:04 |
| clayg | notmyname: yeah that was me doing something stupid with like... monty's patchset 1 or something, i did the rebase right this time and it builds w/o errors everytime - sorry | 18:05 |
| notmyname | clayg: no worries. thanks for looking | 18:05 |
| notmyname | clayg: can you +2/+1 one it (are you logged in to gerrit?) if not, I'll push the button | 18:06 |
| *** wasmum is now known as zz_wasmum | 18:07 | |
| *** AdnanCloud1 has left #openstack-swift | 18:08 | |
| *** kenhui has quit IRC | 18:16 | |
| *** kenhui has joined #openstack-swift | 18:17 | |
| *** openstackgerrit has quit IRC | 18:20 | |
| *** mlipchuk has joined #openstack-swift | 18:28 | |
| *** shri1 has joined #openstack-swift | 18:33 | |
| *** kenhui has quit IRC | 18:33 | |
| *** shri has quit IRC | 18:34 | |
| *** zz_wasmum is now known as wasmum | 18:35 | |
| *** pconstantine_ has joined #openstack-swift | 18:37 | |
| *** openstackgerrit has joined #openstack-swift | 18:38 | |
| *** openstackstatus has joined #openstack-swift | 18:43 | |
| *** ChanServ sets mode: +v openstackstatus | 18:43 | |
| *** shri1 has quit IRC | 18:46 | |
| *** mlipchuk has quit IRC | 18:46 | |
| *** nsquare has quit IRC | 18:46 | |
| *** lpabon has quit IRC | 18:46 | |
| *** tdasilva has quit IRC | 18:46 | |
| *** redbo has quit IRC | 18:46 | |
| *** charz has quit IRC | 18:46 | |
| *** MooingLemur has quit IRC | 18:46 | |
| *** annegentle has quit IRC | 18:46 | |
| *** infotection has quit IRC | 18:46 | |
| *** ekarlso has quit IRC | 18:46 | |
| *** chandan_kumar has quit IRC | 18:46 | |
| *** bkopilov has quit IRC | 18:46 | |
| *** miqui has quit IRC | 18:46 | |
| *** swat30 has quit IRC | 18:46 | |
| *** j_king has quit IRC | 18:46 | |
| *** mitz has quit IRC | 18:46 | |
| *** yuan has quit IRC | 18:46 | |
| *** zhiyan_ has quit IRC | 18:46 | |
| *** wer has quit IRC | 18:46 | |
| *** torgomatic has quit IRC | 18:46 | |
| *** hurricanerix has quit IRC | 18:46 | |
| *** jd__ has quit IRC | 18:46 | |
| *** portante has quit IRC | 18:46 | |
| *** briancline has quit IRC | 18:46 | |
| *** openstackgerrit has quit IRC | 18:46 | |
| *** mhu has quit IRC | 18:46 | |
| *** zaitcev has quit IRC | 18:46 | |
| *** goodes has quit IRC | 18:46 | |
| *** occupant has quit IRC | 18:46 | |
| *** DisneyRicky has quit IRC | 18:46 | |
| *** pandemicsyn has quit IRC | 18:46 | |
| *** mlanner has quit IRC | 18:46 | |
| *** zacksh has quit IRC | 18:46 | |
| *** bgmccollum has quit IRC | 18:46 | |
| *** zigo has quit IRC | 18:46 | |
| *** notmyname has quit IRC | 18:46 | |
| *** anderstj has quit IRC | 18:46 | |
| *** joearnold has quit IRC | 18:46 | |
| *** glange has quit IRC | 18:46 | |
| *** pconstantine_ is now known as pconstantine | 18:46 | |
| *** pconstantine is now known as 18WAABF3U | 18:46 | |
| *** openstackgerrit has joined #openstack-swift | 18:46 | |
| *** shri1 has joined #openstack-swift | 18:46 | |
| *** mlipchuk has joined #openstack-swift | 18:46 | |
| *** nsquare has joined #openstack-swift | 18:46 | |
| *** chandan_kumar has joined #openstack-swift | 18:46 | |
| *** lpabon has joined #openstack-swift | 18:46 | |
| *** bkopilov has joined #openstack-swift | 18:46 | |
| *** mhu has joined #openstack-swift | 18:46 | |
| *** zaitcev has joined #openstack-swift | 18:46 | |
| *** miqui has joined #openstack-swift | 18:46 | |
| *** tdasilva has joined #openstack-swift | 18:46 | |
| *** goodes has joined #openstack-swift | 18:46 | |
| *** swat30 has joined #openstack-swift | 18:46 | |
| *** redbo has joined #openstack-swift | 18:46 | |
| *** charz has joined #openstack-swift | 18:46 | |
| *** occupant has joined #openstack-swift | 18:46 | |
| *** j_king has joined #openstack-swift | 18:46 | |
| *** MooingLemur has joined #openstack-swift | 18:46 | |
| *** mitz has joined #openstack-swift | 18:46 | |
| *** yuan has joined #openstack-swift | 18:46 | |
| *** annegentle has joined #openstack-swift | 18:46 | |
| *** infotection has joined #openstack-swift | 18:46 | |
| *** DisneyRicky has joined #openstack-swift | 18:46 | |
| *** zhiyan_ has joined #openstack-swift | 18:46 | |
| *** ekarlso has joined #openstack-swift | 18:46 | |
| *** wer has joined #openstack-swift | 18:46 | |
| *** torgomatic has joined #openstack-swift | 18:46 | |
| *** pandemicsyn has joined #openstack-swift | 18:46 | |
| *** hurricanerix has joined #openstack-swift | 18:46 | |
| *** mlanner has joined #openstack-swift | 18:46 | |
| *** jd__ has joined #openstack-swift | 18:46 | |
| *** portante has joined #openstack-swift | 18:46 | |
| *** dickson.freenode.net sets mode: +vvvv zaitcev redbo torgomatic portante | 18:46 | |
| *** briancline has joined #openstack-swift | 18:46 | |
| *** joearnold has joined #openstack-swift | 18:46 | |
| *** glange has joined #openstack-swift | 18:46 | |
| *** anderstj has joined #openstack-swift | 18:46 | |
| *** notmyname has joined #openstack-swift | 18:46 | |
| *** zigo has joined #openstack-swift | 18:46 | |
| *** bgmccollum has joined #openstack-swift | 18:46 | |
| *** zacksh has joined #openstack-swift | 18:46 | |
| *** dickson.freenode.net sets mode: +vv glange notmyname | 18:46 | |
| *** mitz has quit IRC | 18:46 | |
| *** mitz- has joined #openstack-swift | 18:47 | |
| *** cschwede has quit IRC | 18:48 | |
| *** tanee has quit IRC | 18:48 | |
| *** dfg has quit IRC | 18:48 | |
| *** jeblair has quit IRC | 18:48 | |
| *** wkelly has quit IRC | 18:48 | |
| *** anticw has quit IRC | 18:48 | |
| *** zackmdavis has quit IRC | 18:48 | |
| *** creiht has quit IRC | 18:48 | |
| *** amandap has quit IRC | 18:48 | |
| *** mtreinish has quit IRC | 18:48 | |
| *** creiht has joined #openstack-swift | 18:48 | |
| *** ChanServ sets mode: +v creiht | 18:48 | |
| *** anticw has joined #openstack-swift | 18:48 | |
| *** dfg has joined #openstack-swift | 18:48 | |
| *** jeblair has joined #openstack-swift | 18:48 | |
| *** cschwede has joined #openstack-swift | 18:48 | |
| *** tanee has joined #openstack-swift | 18:48 | |
| *** amandap has joined #openstack-swift | 18:48 | |
| *** wkelly has joined #openstack-swift | 18:49 | |
| *** zackmdavis has joined #openstack-swift | 18:50 | |
| *** mtreinish has joined #openstack-swift | 18:50 | |
| *** openstackgerrit has quit IRC | 18:51 | |
| *** redbo_ has joined #openstack-swift | 18:51 | |
| *** openstackgerrit has joined #openstack-swift | 18:51 | |
| *** MooingLe1ur has joined #openstack-swift | 18:51 | |
| *** kevinc_ has quit IRC | 18:54 | |
| *** shri1 has quit IRC | 18:56 | |
| *** mlipchuk has quit IRC | 18:56 | |
| *** nsquare has quit IRC | 18:56 | |
| *** lpabon has quit IRC | 18:56 | |
| *** tdasilva has quit IRC | 18:56 | |
| *** redbo has quit IRC | 18:56 | |
| *** charz has quit IRC | 18:56 | |
| *** MooingLemur has quit IRC | 18:56 | |
| *** annegentle has quit IRC | 18:56 | |
| *** infotection has quit IRC | 18:56 | |
| *** ekarlso has quit IRC | 18:56 | |
| *** MooingLe1ur is now known as MooingLemur | 18:57 | |
| *** tdasilva has joined #openstack-swift | 19:00 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift-specs: fix missing link for py-modules on index https://review.openstack.org/102913 | 19:03 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift-specs: Updating containers on object fast-POST https://review.openstack.org/102592 | 19:03 |
| *** kenhui has joined #openstack-swift | 19:04 | |
| *** rook_ has quit IRC | 19:08 | |
| *** kenhui has quit IRC | 19:11 | |
| *** shri1 has joined #openstack-swift | 19:19 | |
| *** mlipchuk has joined #openstack-swift | 19:19 | |
| *** nsquare has joined #openstack-swift | 19:19 | |
| *** lpabon has joined #openstack-swift | 19:19 | |
| *** charz has joined #openstack-swift | 19:19 | |
| *** annegentle has joined #openstack-swift | 19:19 | |
| *** infotection has joined #openstack-swift | 19:19 | |
| *** ekarlso has joined #openstack-swift | 19:19 | |
| *** joeljwright has joined #openstack-swift | 19:24 | |
| *** kevinc_ has joined #openstack-swift | 19:24 | |
| anticw | so, in debugging hell i noticed something else ... if i delete an object ... and then check the container db --- the action there is not reflected | 19:27 |
| anticw | the object itself looks deleted, there is a tombstone .... but the row in the container is untouched | 19:27 |
| anticw | a second delect, which gives 204 from the container server does what's expected ... and in this case the object server sees a tombstone and returns 404 not 20x ... which the proxy reflects | 19:28 |
| notmyname | anticw: did you find a corresponding async pending file? | 19:28 |
| anticw | there isn't one | 19:28 |
| anticw | 100% sure of it | 19:28 |
| anticw | this is 1.13.1 from RDO | 19:28 |
| anticw | it might be zaitcev ruined it :) | 19:29 |
| notmyname | anticw: just performed a PUT/DELETE/listing series on my box. works properly from what I can see | 19:31 |
| notmyname | tombstone is created and it's not in the listing results anymore | 19:31 |
| anticw | notmyname: did you check the deleted flag in the db? | 19:31 |
| anticw | actually, i didn't list ... i just did a delete then checked the db from another process (which is what container-sync is) | 19:32 |
| anticw | let me put a list in there in case it's doing some 'flushing' in some unexpected way | 19:32 |
| anticw | yes ... doing a list causes it to flush | 19:32 |
| anticw | so it's a coherency issue with the sqlite3 code | 19:32 |
| anticw | and if i stop/kill the swift process it seems to not update the flag | 19:33 |
| anticw | let me replace the list with a container-server restart ... arguably that should cause anything pending to get flushed | 19:33 |
| anticw | a process restart won't update the db (from another process PoV) but restart + list does | 19:35 |
| anticw | this means if you do deletes, etc ... and then don't list from the container server, other processes won't see those most recent changes | 19:35 |
| anticw | the swift cli is super chatty and so are a lot of clients, i suspect they mask this in many cases | 19:36 |
| notmyname | anticw: are you talking about the .pending file getting folded in? | 19:36 |
| anticw | yes | 19:36 |
| anticw | .pending is updated | 19:36 |
| anticw | not the db | 19:36 |
| anticw | the issue is when is .pending 'flushed' to the db | 19:36 |
| *** kevinc_ has quit IRC | 19:37 | |
| anticw | ... why this matters is container-sync doesn't see the pending state | 19:37 |
| *** wasmum is now known as zz_wasmum | 19:38 | |
| notmyname | all updates to the db are put into the pending file. nothing is directly put in the db (normal read/write paths). .pending is flushed on some normal read/write paths, but is also update by a background daemon (checking which one...) | 19:38 |
| *** annegent_ has joined #openstack-swift | 19:39 | |
| notmyname | probably the container-updator. still following code paths | 19:39 |
| notmyname | s/or/er/ | 19:39 |
| *** kenhui has joined #openstack-swift | 19:39 | |
| anticw | yeah, i see if i run the updater before container-sync it reacts | 19:40 |
| anticw | i though .pending was only used if the db was locked? | 19:41 |
| anticw | did that change? | 19:41 |
| notmyname | well, lots of things end up calling .merge_items(). listings, puts, reclaim, is_deleted checks, etc | 19:41 |
| notmyname | anticw: nope. been that way for a long time. all writes are written to the .pending record, then they are merged as a group | 19:41 |
| anticw | makes sense ... but i'm sure in the past db updates were more timely other than db locks | 19:43 |
| *** kenhui has quit IRC | 19:43 | |
| *** kenhui has joined #openstack-swift | 19:49 | |
| openstackgerrit | Nirmal Thacker proposed a change to openstack/swift: If the devices path configured in container-server.conf contains a file then an uncaught exception is seen in the logs. For example if file foo exists as such /srv/1/node/foo then when the container-auditor runs, the exception that foo/containers is not a https://review.openstack.org/97148 | 19:51 |
| *** kevinc_ has joined #openstack-swift | 19:52 | |
| *** mmcardle has quit IRC | 19:52 | |
| *** ChanServ sets mode: +v dfg | 19:53 | |
| *** pberis has joined #openstack-swift | 19:56 | |
| *** mlipchuk has left #openstack-swift | 19:58 | |
| *** kenhui has quit IRC | 20:04 | |
| notmyname | just found a new Swift user: http://www.razyr.net (see https://twitter.com/RazyrOps/status/482250873658040321) | 20:06 |
| *** tdasilva has quit IRC | 20:28 | |
| *** zz_wasmum is now known as wasmum | 20:34 | |
| anticw | http://pastebin.com/p8sBfmMF | 20:37 |
| anticw | has me scratching my head | 20:37 |
| anticw | couple of issues, first is swift.common.internal_client.delete_oebject barfs had on 404s ... the second is that when it does we get urllib2.HTTPError not .ClientException ... the third is that I can't catch urllib2.HTTPError | 20:38 |
| openstackgerrit | Joel Wright proposed a change to openstack/python-swiftclient: Add importable SwiftService incorporating shell.py logic https://review.openstack.org/85453 | 20:39 |
| *** wasmum is now known as zz_wasmum | 20:46 | |
| *** tongli has quit IRC | 20:54 | |
| *** pberis has quit IRC | 20:57 | |
| *** mmcardle has joined #openstack-swift | 20:59 | |
| *** cath6184 has joined #openstack-swift | 21:01 | |
| *** DisneyRicky has quit IRC | 21:03 | |
| *** DisneyRicky has joined #openstack-swift | 21:15 | |
| *** nthacker has joined #openstack-swift | 21:20 | |
| *** mmcardle1 has joined #openstack-swift | 21:21 | |
| *** mmcardle has quit IRC | 21:23 | |
| *** cath6184 has quit IRC | 21:24 | |
| *** annegent_ has quit IRC | 21:35 | |
| clayg | anyone know how to use keystone? I have some swiftusertestX users, and some swifttenanttestX teants - but i'm not sure which user is in which tenant | 21:36 |
| *** joeljwright has left #openstack-swift | 21:36 | |
| clayg | I *thought* user-role-list would tell me the answer, it seems to join the demo user to the demo tenant - but it's comin' up empty for these other swiftusertestX users | 21:37 |
| clayg | oh hold the phone now, user-role-list seems to like me better when I pass --user AND --tenant | 21:39 |
| clayg | oh yeah, looks like it's my test:tester test2:tester2 test:tester3 accounts :D | 21:40 |
| clayg | BAM! swift stat -v --os-username swiftusertest1 --os-tenant-name swifttenanttest1 --os-password testing | 21:41 |
| *** kenhui has joined #openstack-swift | 21:45 | |
| *** shri has joined #openstack-swift | 21:46 | |
| *** shri1 has quit IRC | 21:46 | |
| *** redbo_ has quit IRC | 21:49 | |
| *** redbo has joined #openstack-swift | 21:49 | |
| *** ChanServ sets mode: +v redbo | 21:49 | |
| openstackgerrit | Nirmal Thacker proposed a change to openstack/swift: If the devices path configured in container-server.conf contains a file then an uncaught exception is seen in the logs. For example if file foo exists as such /srv/1/node/foo then when the container-auditor runs, the exception that foo/containers is not a https://review.openstack.org/97148 | 21:55 |
| *** mmcardle1 has quit IRC | 22:02 | |
| clayg | acoles: well, i tried :'( | 22:02 |
| clayg | notmyname: do we need to file a bug or something wrt to anne's email on the undoced options in swift? i.e. http://docs.openstack.org/trunk/config-reference/content/container-sync-realms-configuration.html | 22:05 |
| notmyname | clayg: I must have missed that email. but if there are config options that aren't doc'd, then that should be a bug in LP to track it, I think. probably affecting both swift and docs | 22:06 |
| notmyname | annegentle: ^^?? | 22:07 |
| notmyname | in case you missed it (like I did), the -infra team has changed the base image for all the gate jobs. it's now running ubuntu 12.04. the mailing list subject from 2 days ago was "Tempest testing for master is now running on Trusty" | 22:13 |
| torgomatic | so does this mean Swift is no longer tested on Precise or Lucid? | 22:19 |
| *** kenhui has quit IRC | 22:20 | |
| clayg | torgomatic: i test it on precise, lucid sometimes even | 22:22 |
| clayg | torgomatic: oh and cent sometimes too! | 22:22 |
| notmyname | torgomatic: sortof. stable releases (havana and icehouse) are still tested on precise. and oy26 tests are done on ceton6 | 22:22 |
| torgomatic | clayg: Sure; my all-in-one is precise right now. I'm just wondering about the gate stuff. | 22:22 |
| notmyname | but for normal new code, trusty is the thing now | 22:22 |
| clayg | yay the new thing! | 22:22 |
| notmyname | s/yay// | 22:23 |
| *** chandan_kumar has quit IRC | 22:24 | |
| notmyname | "we are doing this because openstack TC has said we will target latest red hat and ubuntu and not break latest LTS" <-- so any change in this practice will need to involve long-term discussions with the TC | 22:30 |
| mattoliverau | Morning | 22:31 |
| notmyname | mattoliverau: hi | 22:31 |
| mattoliverau | notmyname: I watched your PTL presentation last night, nice work :) Must be wierd when your not in control of the slides. | 22:33 |
| notmyname | mattoliverau: ya, it's always been weird doing web-based presentations. the slide control and not seeing the audience are both pretty challenging, It hink | 22:34 |
| *** annegent_ has joined #openstack-swift | 22:36 | |
| mattoliverau | notmyname: I bet, well you did well. Mikal has his today, I think, I should go and heckle :P | 22:36 |
| notmyname | :-) | 22:37 |
| *** annegent_ has quit IRC | 22:42 | |
| openstackgerrit | Matthew Oliver proposed a change to openstack/swift: When a filesystem does't support xattr return a 507 https://review.openstack.org/99883 | 22:42 |
| *** kenhui has joined #openstack-swift | 22:45 | |
| *** annegent_ has joined #openstack-swift | 22:46 | |
| *** kenhui has quit IRC | 22:46 | |
| *** kenhui has joined #openstack-swift | 22:47 | |
| *** cm2 has quit IRC | 22:49 | |
| *** kenhui has quit IRC | 22:49 | |
| *** dmsimard is now known as dmsimard_away | 22:51 | |
| *** cm2 has joined #openstack-swift | 22:53 | |
| zaitcev | mattoliverau: did you test the xattr thing 99883 on a real cluster using a wrong filesystem? | 22:57 |
| zaitcev | mattoliverau: I'm curious if _failsafe_read_metadata plugs your exceptions | 22:58 |
| *** kevinc_ has quit IRC | 23:00 | |
| *** annegent_ has quit IRC | 23:09 | |
| openstackgerrit | Samuel Merritt proposed a change to openstack/swift: Zero-copy object-server GET responses with splice() https://review.openstack.org/102609 | 23:09 |
| *** rpedde has quit IRC | 23:11 | |
| *** zz_wasmum has quit IRC | 23:11 | |
| *** rpedde has joined #openstack-swift | 23:11 | |
| *** zz_wasmum has joined #openstack-swift | 23:32 | |
| *** cm2 has quit IRC | 23:32 | |
| mattoliverau | zaitcev: Sorry was tied up in a meeting.. I tested it briefly on my SAIO, and it logged 'ERROR:root:Filesystem does not support xattr...' to the storageN.error log. But it may need more testing, now that there are storage policies I can change replicas and see if a can get a 507 because qorum wont be reached. | 23:33 |
| torgomatic | you'll never get a 507 from the object server due to quorum; object servers don't know about quorums | 23:34 |
| *** kevinc_ has joined #openstack-swift | 23:34 | |
| mattoliverau | well there you go, well it errors to the log file then. | 23:34 |
| *** eglynn has quit IRC | 23:42 | |
| *** mwstorer has quit IRC | 23:46 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!